Wi-Fi and WiMAX

Solutions
White Paper
Understanding Wi-Fi and WiMAX
as Metro-Access Solutions
This paper provides incumbent wireless Internet service providers (WISPs), new WISPs and
demanding new markets (such as government and education) with a technical analysis of
alternatives for implementing last-mile wireless broadband services.

Contents
Introduction ........................................................................3
Challenges...................................................................3
Wireless Technology Usage Segments ...............................4
Wireless Personal Area Networks ................................5
Wireless Local Area Networks .....................................5
Metropolitan Area and Wide Area Networks ................5
Wi-Fi as a Metro-Access Deployment Option .....................6
Network-Contention Protocol ......................................7
Increasing 802.11 Range Using
Directional Antennas....................................................7
Mesh Networking................................................................7
Lower Initial Costs .......................................................8
Balanced Traffic ...........................................................8
Robustness and Resiliency..........................................8
Spatial Reuse...............................................................8
Network Contention.....................................................9
Benefits .......................................................................9
Challenges...................................................................9
Summary of Wi-Fi Mesh Usage ...................................9
WiMAX as a Metro-Access Deployment Option................10
Fixed..........................................................................10
Portable .....................................................................10
Range and Scalability ................................................11
Flexible Channel Bandwidth ......................................11
Smart Antenna Support .............................................12
Benefits and Challenges ............................................12
Conclusion........................................................................12
Deployment Examples......................................................14
WiMAX Recommended..............................................14
Last Mile .............................................................14
Backhaul .............................................................14
Wi-Fi Mesh Recommended .......................................14
Vertical (Government and Education) ..................14
Key Terminology ...............................................................15
Additional Resources........................................................16
White Papers .............................................................16
General Resources ....................................................16
Summary
This paper provides incumbent wireless Internet
service providers (WISPs), new WISPs and
demanding new markets (such as government
and education) with a technical analysis of
alternatives for implementing last-mile wireless
broadband services. The paper compares current
802.11 single hop and 802.11 mesh networks
with 802.16, a new technology that solves many
of the difficulties in last-mile implementations.
In this paper, the term WISP encompasses Internet
service providers and new market players in vertical
segments (such as government and education) that
want to expand public and private network access
through unlicensed wireless bands.
The paper explains in detail the technical
differences between revisions of 802.11 and
802.16, describes the technology positioning,
and reviews the standards and technology
associated with various usage models. Several
wireless broadband deployment solutions,
scenarios and their merits, including their
challenges, are also described.
2
Introduction
WISPs have been striving for wireless technologies that make
wireless metro access possible. Access to areas that are too
remote, too difcult or too expensive to reach with traditional
wired infrastructures (such as ber) require new technologies
and a different approach.
The three key deployment types that make up wireless metro
access are backhaul, last-mile and large-area coverage (referred
to as hot zones). Wireless last-mile coverage typically uses the
Institute of Electrical and Electronics Engineers (IEEE) 802.11
standard with high-gain antennas, while hot zones use modied
IEEE 802.11 equipment in a mesh deployment.
Open standard radio technologiesincluding 802.11, 802.16
and future standardsoffer advantages to WISPs and users.
For the rst time, industry-wide support and innovation are
driving broadband wireless networking technologies. Network
operators, service providers and users benet from a wide array
of high-performance, feature-rich and cost effective products.
Wireless Fidelity (Wi-Fi) revolutionized the market for
unlicensed client-access radios in a wide variety of
applications. Starting in 2005, Worldwide Interoperability for
Microwave Access (WiMAX) certication of the IEEE 802.16-
2004 standard for xed-position radios will do the same for
point-to-point (P2P) and point-to-multi-point (P2MP) wireless
broadband equipment in both the licensed and unlicensed
bands. In 2006, the IEEE 802.16e standard for portable
operation is expected to be ratied, thus standardizing client
radios in unlicensed and licensed bands. This certication will
provide users with an alternative and allow service providers
the benet of additional tier services.
The cost and limited flexibility of wired backhaul limits wireless
access growth. In the face of the technical challenges, WISPs
have begun to look ahead at WiMAX-certified solutions, which
will be available in early 2005.
To date, WISPs have capitalized on the cost and complexity
associated with traditional high-speed wired broadband
infrastructures by applying ingenuity to solve last-mile
problems. WISPs modied existing wireless technologies,
typically based on the IEEE 802.11 standard, to patch last-mile
gaps. Limitations in these deployments surfaced, however.
Because wired backhaul solutions can be too expensive for
establishing widespread wireless access and because a
standard means for deploying IEEE 802.11 into the last mile or
within a hot zone has not emerged, each WISP implements
long-distance IEEE 802.11 solutions differently.
WiMAX is a wireless metropolitan-area network technology
that provides interoperable broadband wireless connectivity
to fixed, portable and nomadic users. It provides up to 50-
kilometers of service area, allows users to get broadband
connectivity without the need of direct line-of-sight to the
base station, and provides total data rates up to 75 Mbps
enough bandwidth to simultaneously support hundreds of
businesses and homes with a single base station.
This white paper discusses wireless metro-access
technologies: Wi-Fi with high gain antennas, Wi-Fi meshed
networks and WiMAX. It explores how the technologies differ
and how they can be combined to provide a total last-mile
access solution now and in the future.
Challenges
Typical modied IEEE 802.11 network topologies associated
with last-mile and hot-zone coverage use either directional
antennas or a mesh-network topology. Wi-Fi provides the
certication for IEEE 802.11 client-to-access point (AP)
communications. However, implementations of AP-to-AP and
AP-to-service providers (that is, backhaul applications) that are
typically needed for wireless last-mile and hot-zone coverage
are still proprietary, thus providing little or no interoperability.
Because the IEEE 802.11 standards were designed for
unwiring the local area network (LAN), metro-access
applications are facing the following challenges:
Non-standard wireless inter-AP communication.
Today, wireless links used to connect 802.11 APs for inter-AP
communication in mesh networking are vendor-specic. The
proposed IEEE 802.11s standard, estimated to be ratied in
2007, will standardize Wi-Fi mesh networking.
Providing quality of service (QoS). QoS refers to the
ability of the network to provide better service to selected
network traffic over various technologies. The goal of QoS
technologies is to provide priority (including dedicated
bandwidth to control jitter and latency) that is required by
some real-time and interactive traffic, while making sure that
in so doing the traffic on the other paths does not fail. In
general, unlicensed bands can be subject to QoS issues
because deployment is open to anyone. Advances in the
associated standards and related technologies, however,
help mitigate problems with unlicensed bands, such as
multi-path interference. The proposed IEEE 802.11e
standard, which is projected to be ratified in 2006, will
standardize Wi-Fi mesh-network topology.
3
Expensive backhaul costs. Backhaul refers both to the
connection from the AP back to the provider and to the
connection from the provider to the core network. To extend
wireless access nodes, providers still rely on wires for long-
distance coverage. Some providers find wiring large areas
too expensive.
Limited services. Without QoS, applications such as
voice over Internet protocol (VoIP) may reduce a calls
quality, thus limiting the providers ability to tier services and
obtain additional revenue streams. Current Wi-Fi last-mile
and large-coverage solutions offer excellent data transfers.
Some vendors offer proprietary QoS.
Despite the challenges, wireless metro-access solutions are
continuously sought after for the following reasons:
Wireless metro-access solutions available today, such as
mesh networking implementations, are more cost-effective
and flexible than their wired counterparts.
These solutions provide a standards-based connection from
AP-to-mobile users for hot-zone coverage.
WISPs can offer broadband services to geographically
challenged areas (such as rural towns).
Local governments can provide free access for businesses
or emergency services (such as police and fire fighters).
Educational institutions can broaden learning through online
collaboration between students and faculty on and off campus.
Enterprises and large private networks can communicate
and monitor supply-chain activities in near real time.
Wireless Technology
Usage Segments
The reasons behind wireless deployments are as diverse as
the wireless technologies being offered today. Each wireless
technology is designed to serve a specific usage segment:
Personal area networks (PANs)
Local area networks (LANs)
Metropolitan area networks (MANs)
Wide area networks (WANs)
The requirements for each usage segment are based on a
variety of variables, including:
Bandwidth needs
Distance needs
Power
User location
Services offered
Network ownership
Optimized applications exist for each usage segment. For
example, in some locations it is possible to seamlessly use a
third-generation handset while traveling country to country
while in a wireless WAN environment.
Figure 1 shows the wireless standards organizations, the
standards, and their capabilities (bandwidth and distance)
mapped to the four usage segments previously mentioned.
Figure 1. Wireless technologies target segments.
The three standards organizations in Figure 1 are:
Institute of Electrical and Electronics Engineers (IEEE)
European Telecommunications Standards Institute (ETSI)
Third-Generation Partnership Project (3GPP)
The IEEE and ETSI standards are interoperable and focus primarily
on wireless packet-based networking. The 3GPP standard
focuses on cellular and third-generation mobile systems.
Each usage segment has a corresponding wireless standard,
but segment overlaps do exist. For example, ultra-wide band
(UWB) supports faster le transfers and could allow a user to
transport les faster than when using Wi-Fi or WiMAX. A user
might employ UWB to quickly transfer a le, but distance
limitations prevent this from becoming common place in the LAN
or MAN. In the case of Wi-Fi and WiMAX, the lines may seem
even more unclear. WISPs deploying last-mile solutions with Wi-
Fi and directional antennas or using a Wi-Fi mesh topology over
large areas may make it seem as though Wi-Fi is moving into the
MAN. However, understanding usage segments and the
associated technologies make it apparent that very little overlap
exists. These technologies are extremely different, as can be
shown by reviewing the three different implementation types.
4
5
1. Wi-Fi with directional antennasFocus on patching
last-mile gaps. Currently, WISPs offering alternatives to
Digital Subscriber Line (DSL) and Data Over Cable
Interface Specication (DOCSIS) use this solution.
However, long-range limitations exist, specically in dealing
with multi-path interference. In addition, QoS and
proprietary equipment reduce scalability. These
approaches are looking ahead to WiMAX to ll in the last
mile, to offer QoS and to support backhaul.
2. Wi-Fi with a mesh-network topologyWISPs use this
technology to cover large areas and to extend the reach of
the LAN for both indoor and outdoor applications. A mesh-
networking topology provides a more scalable solution than
a solely directional-antenna implementation. More APs (also
referred to as nodes) are deployed in dense clusters to get
around wire dependency and to increase client coverage.
However, QoS and inter-AP communications are still
proprietaryWISPs using this technology are looking
ahead to WiMAX to provide inter-AP backhaul and point-of-
presence (PoP) backhaul.
3. WiMAX802.16-2004-based devices offer WISPs a
standards-based long-range, higher performance solution.
WiMAX devices for the last mile will not be available until 2005.
This new technology offers a great deal of potential.
Wireless Personal Area Networks
The first usage segment is the wireless PAN, shown in the left
column of Figure 1. Typical network coverage in the PAN is up
to ten meters, but performance varies depending on the
standard employed.
The IEEE 802.15.1 standard (also called Bluetooth) is primarily
used for unwiring computing and communication peripherals,
such as a computer to a printer or a handset to a headset.
Bluetooth is rated up to 1 Mbps in performance data rates.
The second standard in this usage segment, the IEEE 802.15.3
standard (also called ultra-wide band), is designed for
delivering multimedia services. UWB supports data rates over
400 Mbps, allowing for video of digital video disc (DVD) quality
to be shared throughout the home. In this case the PAN
becomes a high-speed personnel area network.
Wireless Local Area Networks
Figure 1 shows WLANs in the second column. Standards-
based WLANs typically:
Service more applications and users than do PANs
Cover a greater distance than PANs: up to 100 meters
Aggregate PANs
The wireless standard associated with WLANs is IEEE 802.11.
Three major revisions to the physical layer have been released:
802.11a supports bandwidth speeds up to 54 Mbps
802.11b supports bandwidth speeds up to 11 Mbps
802.11g supports bandwidth speeds up to 54 Mbps
WISPs using directional antennas or implementing pre-
standard Wi-Fi mesh topologies have been able to increase
performance beyond 54 Mbps and to cover over ten
kilometers (km) in range using the 802.11 standard. The
increase in range has placed 802.11 into two usage
segments: LAN and MAN. This overlap is shown in Figure 1.
Range alone does not constitute a usage segment because
the segments are based on a variety of variables, such as the
distance between APs and the number of users.
Metropolitan Area and Wide Area Networks
The wireless MAN is the third usage segment shown in Figure 1.
The wireless MAN aggregates LANs and typically covers areas up
to 50 km. Both WiMAX and coppered wired technologies (such as
DSL and DOCSIS cable) are used in this usage segment.
The fourth usage segment shown in Figure 1 is the WAN.
WANs aggregate MANs across large geographic areas (over
50 km). WAN uses a variety of communication media to pass
large amounts of trafc from the various MANs. However, the
most common media used are ber optic links. This set of
high-speed, high-bandwidth interconnections is referred to as
the core network. Performance of WAN networks is up to 10
Gbps and depends on the type of trafc the network handles:
voice only or voice, video and data.
Today, the popularity, cost benefits and throughput
associated with Wi-Fi networks have caused a growth in
network deployments, use and adoption. This is due to the
options available in achieving access to the last mile. WISPs
are pushing Wi-Fi to the limits to reach and cover MANs.
In 2005, WISPs have additional options with the implementation
of WiMAX, which copes with some of the challenges facing Wi-Fi.
Examining the technical capabilities of the three options
available for WISPs to cover the last mile provides a deeper
understanding of each of these options.
Wi-Fi as a Metro-Access
Deployment Option
The Wi-Fi certification addresses interoperability across
IEEE 802.11 standards-based products. The IEEE 802.11
standard, with specific revisions, was designed to address
wireless local area coverage.
External modifications to the standard through hardware and
software allow Wi-Fi products to become a metro-access
deployment option. These two major modifications address
two different usage models:
Fixed-access or last-mile usage802.11 with high-
gain antennas
Portable-access or hot-zone usage802.11
mesh networks
Wi-Fi products associated with the metro-access deployment
option use these different radio frequencies:
The 802.11a standard uses 5 GHz in an AP-to-AP interlink.
The 802.11b and 802.11g standards use 2.4 GHz.
The 802.11a, 802.11b and 802.11g standards use different
frequency bands; devices based on these standards do not
interfere with one another. On the other hand, devices on
different bands cannot communicate; for example, an
802.11a radio cannot talk to an 802.11b radio.
The most common deployments by WISPs for wireless metro
access to date are the 802.11b and 802.11g standards
because of interoperability and the greater range they achieve
in the 2.4-GHz band.
Each standard also differs in the type of radio-modulation
technology used, as follows:
The 802.11b standard uses direct-sequence spread spectrum
(DSSS) and supports bandwidth speeds up to 11 Mbps.
The 802.11a and 802.11g standards use orthogonal
frequency division multiplexing (OFDM) and support speeds
up to 54 Mbps. Because OFDM is more adaptable to
outdoor environments and interference, it is most commonly
used for metro-access solutions.
OFDM technology uses sub-carrier optimization, which assigns
small sub-carriers to users based on radio frequency conditions.
Orthogonal means that the frequencies into which the carrier
is divided are chosen such that the peak of one frequency
coincides with the nulls of the adjacent frequency. The data
stream is converted from serial to parallel, and each parallel
data stream is mapped by a modulation block. The
modulated data is fed to an inverse fast Fourier transform
(IFFT) block for processing. The IFFT block converts the
discrete modulated frequencies into a time-domain signal,
which is used to drive the radio frequency (RF) amplifier.
This enhanced spectral efficiency is a great benefit to OFDM
networks, making them well suited for high-speed data
connections in both fixed and mobile solutions.
The 802.11 standard provides for 64 subcarriers. These
individual carriers are sent from the base station (BS) or AP to
the subscriber station (SS) or client and are then reconstituted
at the client side. In non-line-of-sight (NLOS) situations, these
carriers will hit walls, buildings, trees and other objects, which
then reflect the signal, creating multi-path interference.
By the time the carrier signals reach the client for
reconstitution, the individual carrier signals are time delayed.
For example, one carrier may have been reected once and
arrived 1 s later than another, and a second carrier may have
been reected twice and arrive 2 s later. The larger number of
subcarriers over the same band results in narrower
subcarriers, which is the equivalent to larger OFDM symbol
periods. Consequently, the same percentage of guard time or
cyclic prex (CP) will provide larger absolute values in time for
larger delays, improving resistance to multi-path interference.
Because the 802.11a and 802.11g standards use OFDM, they
are more resilient than the 802.11b standard in outdoor multi-
path-prone environments. These factors were taken into
account when developing the 802.16-2004 standard. The
802.11a and 802.11g standards have one-fourth of the OFDM
symbol options for CP than in the 802.16-2004 standard.
Table 1: Wi-Fi standards at a glance.
Wi-Fi Standard Frequency Modulation
802.11a 5 GHz OFDM
802.11b 2.4 GHz DSSS
802.11g 2.4 GHz OFDM
The 802.11g standard is often selected for a last-mile solution
for three reasons.
Speed
The ability to handle interference
Interoperability with 802.11b-based devices
6
7
Network-Contention Protocol
The 802.11 standard uses a carrier-sense, multiple-
access/collision-avoidance (CSMA/CA) protocol, which is a
network-contention protocol that listens to the network to
avoid transmission collisions.
CSMA/CA contributes to network trafc by controlling the
network. Before real data is transmitted across the network,
CSMA/CA broadcasts a signal onto the network to listen for
collision scenarios and to tell all other devices not to broadcast.
Some equipment manufacturers and WISPs have been able to
get around hidden-node problems by introducing APs or using
intelligent APs to monitor trafc. WiMAX (IEEE 802.16-2004)
uses a scheduling protocol, with all scheduling owned by the
base-station AP, thus improving reliability.
As the number of users on an 802.11 network increases,
however, the efciency of the network decreases because of
the overhead of managing additional subscribers. Whereas a
single user may enjoy over 30 Mbps of throughput in an
802.11g network when only one user is accessing the AP, by
the time 100 users are accessing the AP, the per user
throughput can be less than dial-up rates.
Increasing 802.11 Range Using
Directional Antennas
Figure 2: 802.11 last-mile networks.
Omni-directional antennas are often used for line-of-sight
communications with mobile stations spread out in all
directions. Because it is not necessary to broadcast to the
clouds, omni-directional antennas propagate RF signals in all
directions equally on a horizontal plane (that is, throughout a
facility), but limit range on the vertical plane. Their radiation
pattern resembles that of a large doughnut with the antenna at
the center of the hole. Omni-directional antennas are commonly
used in traditional WLAN settings and mesh networks.
A directional antenna transmits and receives RF energy more
in one direction than others. This radiation pattern is similar to
the light that a flashlight or spotlight produces. Higher-gain
antennas have a narrower beam width, which limits coverage
on the sides of the antennas. Directional antennas typically
have gains much higher than omni-directional antennas.
High-gain antennas work best for covering a large distance in
narrow areas or for supporting point-to-point links between
buildings. In some cases, a directional antenna can reduce
the number of APs needed within a facility. For example, a
long loading dock of a distribution center many require three
APs having omni-directional antennas. But the use of a high-
gain directional antenna in the same situation would only
require a single AP. Using 802.11 with high-gain antenna can
bridge last-mile gaps, but they require more power.
Mesh Networking
Mesh-network topology extends the range of traditional LANs
and WLANs. In a mesh-network topology, each node is
connected and communication protocols are shared across the
nodes. A Wi-Fi mesh infrastructure is formed when a collection
of 802.11a, b, or g-based nodes are interconnected by wireless
802.11 links. The 802.11a standard is most commonly used in
AP-to-AP links because of its performance and non-channel
overlap with 802.11b or 802.11g transmissions. Mesh networks
automatically learn and maintain dynamic path congurations.
Wireless devices in a mesh-network topology create a seamless
path for data to one another over a license-exempt spectrum at
2.4 or 5 GHz, with speeds up to 54 Mbps.
Current backend implementations of Wi-Fi mesh
infrastructures are based on proprietary solutions. These
proprietary-based solutions may support VoIP and QoS.
They can also increase the coverage range of standard Wi-Fis
100-meter limit to over 10 km. In addition, performance can
be increased from the Wi-Fis 54-Mbps limit to over 100 Mbps.
These implementations, however, are not interoperable, have
limited scalability and in certain deployments are limited by
wired backhaul. The ratication of 802.11s will standardize the
Wi-Fi mesh-network topology. The 802.11s standard is
estimated to be ratied in 2007. Wi-Fi mesh-network
topologies can be used as a last-mile solution but are better
suited to blanket large areas with 802.11 access.
Mesh networking is sometimes referred to as multi-hop
networking. Mesh topologies provide a flexible architecture
that can move data between nodes efficiently.
Within the mesh network, small nodes act as simple routers.
The nodes are installed throughout a large area (such as a
neighborhood or a school campus). Each node then transmits
a low power signal capable of reaching neighboring nodes,
each of which in turn transmits the signal to the next node,
with the process being repeated until the data arrives at its
destination. An advantage of this topology is the ability for the
deployment to navigate around a large obstacle, such as a
mountain which can otherwise block a subscriber from
reaching a base station. In a mesh network, blocked
subscribers can get to the base station indirectly by going
through other nodes. Even a small amount of meshing can
greatly improve a base stations coverage if sufficient small
nodes are in place.
Mesh networks provide advantages over direct line-of-sight
implementations because they can adapt to changes in
network topology. Nodes can be readily added and removed,
and their location changed. As people become more mobile
and wireless capabilities are included in new classes of
devices, future business and home networks need to adapt or
self-configure to these changes.
Benefits of the mesh-network topology include lower initial
costs, balanced traffic, mobility and availability.
Lower Initial Costs
Mesh networks provide lower costs to the operator because
users already have a client (such as a laptop with embedded
Wi-Fi technology).
Balanced Trafc
Mesh networks provide greater redundancy and can be used
for traffic balancing. In dense networks, such as crowded
offices or apartments, each device can have many neighbors
creating multiple paths between two communicating devices.
In the presence of localized interference, a multi-hop network
can route data along an alternate path.
If only one node requires a large amount of bandwidth, then
the network can dynamically route traffic to other network
nodes, avoiding the congested node. Current single-hop
networks do not have the ability to dynamically adapt to
interference or overburdened network nodes.
Mesh networks have some advantages over the single-hop
and directional last-mile alternatives, such as:
Robustness
Resiliency
Spatial reuse
Robustness and Resiliency
Mesh-network topologies are more robust than single-hop
networks because they are not dependent on the
performance of a single node for operation. In a single-hop
network, if the node goes down, so does the network. In
mesh-network architecture, if the nearest node goes down or
if localized interference occurs, the network continues to
operate; data is simply routed along an alternate path.
A good example of an application needing both robustness
and resiliency is email, which is divided into data packets that
are sent across the Internet via multiple routes then
reassembled into a coherent message that arrives in the
recipients mailbox. Using multiple routes to deliver data
increases the effective bandwidth of the network.
Spatial Reuse
Mesh networks use available bandwidth efciently. In a single-
hop network, devices must share a node. If several devices
attempt to access the network at once, a trafc jam occurs and
the system slows. By contrast, in a mesh network, many devices
can connect to the network at the same time through different
nodes, without necessarily degrading system performance. The
shorter transmission ranges in a mesh network limit interference,
allowing simultaneous, spatially separated data ows.
To deploy a mesh network cost effectively, however, service
providers need a large initial subscriber base. Mesh networks
are built from close proximity to the point-to-point connection,
then expanding out. Service providers typically do not build a
mesh network until they have established a subscriber base.
8
Figure 3: 802.11 mesh networking.
Mesh networks can create noticeable latency. Even when
individual hops introduce small delays, latency increases with
every hop. Application services that require low latency, such
as VoIP and live video conferences over the Internet, can be
adversely affected. As more nodes are added in the mesh-
network topology, bottlenecks begin to appear as the nodes
closest to the PoP function mostly as aggregators and routers
for the trafc to the edges. This impacts the bandwidth
availability for the subscribers for these nodes. The latency
problem gets progressively worse as new subscribers are
added near the edge.
Mesh networks are also noisy. By necessity (until smart
directional antennas are used at the client end, which is
unlikely for a long time), every node in the mesh-network
topology is an omni-directional broadcaster. This produces
noise, which the nodes detect as errors. Each error forces
retransmissions along the entire chain, with the throughput
degrading progressively as the transmissions aggregate to
the PoP. This reduces the bandwidth for those closest to the
node as well as those farther away.
Both 802.16-2004 and 802.11 at a given power level and a
given channel bandwidth (by default, 20 MHz for 802.11) have
similar ranges. Range is the easiest problem to solve when
attempting to use 802.11 for outdoor applications. 802.16 was
built from the ground up to address last-mile limitations,
however, and feedback from WISPs deploying 802.11 for the
last mile was addressed when developing the 802.16 standard.
Network Contention
The network-contention protocol on which 802.11 is based
can be the root of problems in mesh-network topologies as
well. CSMA/CA can cause a hidden-node problem when
clients are out of reach of one another.
Much as a broadcast storm on a wired LAN segment can
bring trafc to a standstill, hidden transmitter nodes interfering
with one another have a detrimental effect on the performance
of every wireless node in the mesh network. This interference
can cause overall performance of the entire wireless network
to drop dramatically.
The revisions to the 802.11 standard address the hidden-
transmitter problem using special packets called request to send
(RTS), clear to send (CTS) and acknowledge (ACK). The 802.11a,
b and g standards use these special packets to alert every node
on the network that a transmission node has data to send, that a
transmission is about to take place, and that the transmission has
ended by broadcasting these packets across the entire network.
This is a time- and bandwidth-consuming process required for
each and every transmission by each and every wireless node
in the network. In addition, proper AP placement and network
design can help control hidden node issues.
Benets
The benefits of using Wi-Fi for last-mile solutions now are:
Off-the-shelf 802.11 standard products are currently available
Initial investment is cost effective for small deployments
Flexibility over wired installations can be achieved
Challenges
The limitations to mesh networks are:
A large subscriber base is needed to cover larger areas
Shared bandwidth
Latency
Proprietary implementation
Standardized Wi-Fi mesh-network topology wont be
available until the implementation of the 802.11s standard
Standardized Wi-Fi QoS wont be available until the
implementation of the 802.11e standard
Summary of Wi-Fi Mesh Usage
The majority of WISPs and service providers are using Wi-Fi
mesh-network topology for:
Wireless coverage of LANs
Blanketing large areas with hot-zone coverage
9
WiMAX as a Metro-Access
Deployment Option
WiMAX is a worldwide certication addressing interoperability
across IEEE 802.16 standards-based products. The IEEE 802.16
standard with specic revisions addresses two usage models:
Fixed
Portable
Fixed
The IEEE 802.16-2004 standard (which revises and replaces
IEEE 802.16a and 802.16REVd versions) is designed for
fixed-access usage models. This standard may be referred to
as fixed wireless because it uses a mounted antenna at the
subscribers site. The antenna is mounted to a roof or mast,
similar to a satellite television dish. IEEE 802.16-2004 also
addresses indoor installations, in which case it may not be as
robust as in outdoor installations.
The 802.16-2004 standard is a wireless solution for xed
broadband Internet access that provides an interoperable,
carrier-class solution for the last mile. The Intel WiMAX solution
for xed access operates in the licensed 2.5-GHz, 3.5-GHz
and license-exempt 5.8-GHz bands. This technology provides
a wireless alternative to the cable modem, digital subscriber
lines of any type (xDSL), transmit/exchange (Tx/Ex) circuits and
optical carrier level (OC-x) circuits.
Figure 4: WiMAX network topology.
Portable
The IEEE 802.16e standard is an amendment to the 802.16-
2004 base specification and targets the mobile market by
adding portability and the ability for mobile clients with IEEE
802.16e adapters to connect directly to the WiMAX network
to the standard. The 802.16e standard is expected to be
ratified in early 2005.
The 802.16e standard uses Orthogonal Frequency Division
Multiple Access (OFDMA), which is similar to OFDM in that it
divides the carriers into multiple subcarriers. OFDMA, however,
goes a step further by then grouping multiple subcarriers into
sub-channels. A single client or subscriber station might
transmit using all of the sub-channels within the carrier space,
or multiple clients might transmit with each using a portion of
the total number of sub-channels simultaneously.
The IEEE 802.16-2004 standard improves last-mile delivery in
several key aspects:
Multi-path interference
Delay spread
Robustness
Multi-path interference and delay spread improve performance
in situations where there is not a direct line-of-sight path
between the base station and the subscriber station.
The emerging 802.16-2004 media-access control (MAC) is
optimized for long-distance links because it is designed to
tolerate longer delays and delay variations. The 802.16
specification accommodates MAC management messages
that allow the base station to query the subscriber station, but
there is a certain amount of time delay.
WiMAX equipment operating in license-exempt frequency
bands will use time-division duplexing (TDD); equipment
operating in licensed frequency bands will use either TDD or
frequency-division duplexing (FDD). Intel WiMAX products will
support TDD and half-duplex FDD operation.
The IEEE 802.16-2004 standard uses OFDM for optimization
of wireless data services. Systems based on the emerging IEEE
802.16-2004 standards are the only standardized OFDM-
based, wireless metropolitan area networks (WMAN) platforms.
In the case of 802.16-2004, the OFDM signal is divided into
256 carriers instead of 64 as with the 802.11 standard. As
previously stated, the larger number of subcarriers over the
same band results in narrower subcarriers, which is
equivalent to larger symbol periods. The same percentage of
guard time or cyclic prefix (CP) provides larger absolute
values in time for larger delay spread and multi-path immunity.
The 802.11 standard provides one-fourth of the OFDM
options for CP than does the 802.16-2004 standard, which
provides
1
/32,
1
/16,
1
/8 and
1
/4, where each can be optimally
set. For a 20-MHz bandwidth, the difference between a
1
/4
CP in .11 and 16 would be a factor of four because of the
ratio
256
/64. In OFDMA with 2048 FFT size, the ratio is 32.
10
The physical layers (PHYs) for both 802.11 and 802.16-2004
are designed to tolerate delay spread. Because the 802.11
standard was designed for 100 meters, it can tolerate only
about 900 nanoseconds of delay spread. The 802.16-2004
standard tolerates up to 10 microseconds of delay spread
more than 1000 times than in the 802.11 standard.
Range and Scalability
The 802.16-2004 standard relies upon a grant-request access
protocol that, in contrast to the contention-based access used
under 802.11, doesnt allow data collisions and, therefore,
uses the available bandwidth more efciently. No collisions
means no loss of bandwidth due to data retransmission. All
communication is coordinated by the base station. Other
characteristics of the 802.16-2004 standard include:
Improved user connectivityThe 802.16-2004
standard keeps more users connected by virtue of its
flexible channel widths and adaptive modulation. Because it
uses channels narrower than the fixed 20-MHz channels
used in 802.11, the 802.16-2004 standard can serve lower-
data-rate subscribers without wasting bandwidth. When
subscribers encounter noisy conditions or low signal
strength, the adaptive modulation scheme keeps them
connected when they might otherwise be dropped.
Higher quality of serviceThis standard also enables
WISPs to ensure QoS for customers that require it and to
tailor service levels to meet different customer requirements.
For example, the 802.16-2004 standard can guarantee high
bandwidth to business customers or low latency for voice
and video applications, while providing only best-effort and
lower-cost service to residential Internet surfers.
Full support for WMAN serviceFrom its inception, the
802.16-2004 standard was designed to provide WMAN
service. Hence, it is able to support more users and deliver
faster data rates at longer distances than last-mile
implementations based on the 802.11g standard.
Robust carrier-class operationThe standard was
designed for carrier-class operation. As more users join, they
must share the aggregate bandwidth and their individual
throughput decreases linearly. The decrease, however, is
much less dramatic than what is experienced under 802.11.
This capability is termed efcient multiple access.
Flexible Channel Bandwidth
As the distance between a subscriber and the base station (or
AP) increases, or as the subscriber starts to move by walking
or driving in a car, it becomes more of a challenge for that
subscriber to transmit successfully back to the base station at
a given power level. For power-sensitive platforms such as
laptop computers or handheld devices, its often not possible
for them to transmit to the base station over long distances if
the channel bandwidth is wide. The 802.11 channel
bandwidth is fixed at 20 MHz. In contrast, applications
modeled on third-generation principles limit channel
bandwidth to about 1.5 MHz to provide longer range.
The IEEE 802.16-2004 and IEEE 802.16e standards have
flexible channel bandwidths between 1.5 and 20 MHz to
facilitate transmission over longer ranges and to different
types of subscriber platforms. In addition, this flexibility of
channel bandwidth is also crucial for cell planning, especially
in the licensed spectrum. For scalability, an operator with 14
MHz of available spectrum, for example, may divide that
spectrum into four sectors of 3.5 MHz to have multiple
sectors (transmit/receive pairs) on the same base station.
With a dedicated antenna, each sector has the potential to
reach users with more throughput over longer ranges than
can an omni-directional antenna. Net-to-net, flexible channel
bandwidth is imperative for cell planning.
The 802.16-2004 standard has strong commercial backing to
go along with its technical capabilities. The WiMAX Forum*, a
nonprofit group that promotes 802.16-2004 technology, has
as its goal the certification of interoperable 802.16-2004
standard products, regardless of vendor. In that regard, the
forum is following the lead of the Wi-Fi Alliance*, which helped
popularize and commercialize 802.11 standard technology.
Founded in 2003 by wireless service providers and equipment
manufacturers, the WiMAX Forum now includes almost 70
member companies. Several of them expect to deliver
WiMAX-certified* products later this year.
11
12
Smart Antenna Support
Smart antennas are being used to increase the spectral density
(that is, the number of bits that can be communicated over a
given channel in a given time) and to increase the signal-to-
noise ratio for both Wi-Fi and WiMAX solutions. Because of
performance and technology, the 802.16-2004 standard
supports several adaptive smart antenna types, including:
Receive spatial diversity antennasEntails more than
one antenna receiving the signal. The antennas need to be
placed at least half a wavelength apart to operate effectively.
Note that wavelength can be derived by taking the inverse
of the frequency. For example, for a 2.5-GHz carrier the
wavelength would be 0.13 meters or 5.1 inches. For a 5.8-
GHz carrier the wavelength would be 0.05 meter or 1.9
inches. When considering half a wavelength for the
frequencies of interest, we are talking one to two and a half
inches. Maintaining this minimum distance ensures that the
antennas are incoherent, that is, they will be impacted
differently by the additive/subtractive effects of signals
arriving by means of multiple paths.
Simple diversity antennasDetect the signal strength of
the multiple (two or more) antennas attached and switch that
antenna into the receiver. The more incoherent antennas to
choose from, the higher the likelihood of getting a strong signal.
Beam-steering antennasShape the antenna array
pattern to produce high gains in the useful signal direction
or notches that reject interference. High antenna gain
increases the signal, noise and rate. The directional pattern
attenuates the interference out of the main beam. Selective
fading can be mitigated if multi-path components arrive with
a sufficient angular separation.
Beam-forming antennasAllow the area around a base
station to be divided into sectors, allowing additional
frequency reuse among sectors. The number of sectors can
range from as few as four to as many as 24. Base stations
that intelligently manage sectors have been used for a long
time in mobile-service base stations.
Benets and Challenges
The key benefits of WiMAX are:
Built-in QoS
High performance
Standards-based
Smart antenna support
The most signicant challenge is that WiMAX is a new technology
with emerging support.
Conclusion
Wi-Fi mesh networks are driving the demand for WiMAX by
increasing the proliferation of wireless access, increasing the
need for cost-effective backhaul solutions and faster last-mile
performance. WiMAX can be used to aggregate Wi-Fi
networks (such as mesh-network topologies and hotspots)
and Wi-Fi users to the backend. As Figure 5 illustrates, each
wireless metro-access solution has common and unique
benefits. Today, a Wi-Fi mesh-network offers mobility, while
WiMAX offers a long-distance backhaul and last-mile solution.
The best solution is a combination of the two.
Figure 5: 2004-2005 WiMAX and Wi-Fi metro-access solution features.
WISPs have a variety of wireless deployment options for
covering large areas and last-mile gaps. The best solution
varies based on usage models, time of deployment,
geographic location and network application type (such as
data-only, VoIP and video). Each deployment can be tailored to
meet the network needs of its users.
WISPs wanting to deploy wireless solutions before the rst half
of 2005 for mobile users who need connections using laptop
computers over large areas and who do not require QoS can
use Wi-Fi mesh-network topology to extend the reach of
traditional WLANs. Wi-Fi mesh infrastructures are optimized for
data-only usage models, however. QoS add-ons are available
for proprietary Wi-Fi mesh implementations but are not
standardized and limit multi-vendor interoperability.
Intel is working within the wireless industry to drive the
deployment of both Wi-Fi and WiMAX networks. Intel is one of
the founders of the WiMAX Forum, the industry-led, non-prot
corporation formed to promote and certify the compatibility and
interoperability of broadband wireless products. In addition, Intel
has help proliferate Wi-Fi adoption with the Intel

Centrino
mobile technology brand. In 2005, WiMAX and various forms of
Wi-Fi will be combined to offer optimal end-to-end performance.
Wi-Fi WLANs and Mesh will coexist with WiMAX.
Recommendations for deployments:
Use 802.16-2004 for P2P and P2MP links in rural, low-
density areas.
Deploy unlicensed Wi-Fi networks today to capture the
benets of open-standard radio and to bring low-cost wireless
service to urban and suburban areas. If fully licensed operation
is desired for increased RF predictability, introduce licensed
WiMAX 802.16-2004 into metro-infrastructures or, when
available, add WiMAX 802.16e for mobile-client devices.
If unlicensed operation is desired for maximum flexibility at
lowest cost, monitor the price-feature-performance trends
of evolving WiMAX and Wi-Fi versions and introduce WiMAX
if, and when, these trends move in its favor.
In 2005, WISPs wanting QoS, standards-based, scalable
solutions can use WiMAX to cover last-mile deployments.
WiMAX (802.16-2004) provides broadband connectivity to
proprietary and standards-based Wi-Fi mesh networks,
WLANs, hotspots, residences and businesses.
WiMAX (802.16-2004) provides wireless broadband
connectivity to areas beyond the reach of traditional
broadband (such as DOCSIS cable, xDSL and T1) and
enables Wi-Fi mesh-network topology growth.
With attention focused on WiMAX, its easy to forget that Wi-Fi
is also rapidly evolving. Wi-Fi radios are appearing not just in
laptops and personal digital assistants (PDAs), but in
equipment as diverse as mobile phones, parking meters,
security cameras and home entertainment equipment. As a
result of the increasing adoption rate, Wi-Fi will continue to
become faster, more secure, more reliable and more full-
featured. In turn, these advances will drive continued adoption.
A combined Wi-Fi mesh and WiMAX deployment, as shown in
Figure 6, offers a more cost-effective solution than a sole
Wi-Fi directional-antenna deployment or a Wi-Fi mesh
network with wired backhaul for WISPs that want to extend
the LAN or cover the last mile.
Figure 6: Phase 1 WiMAX backhaul for a Wi-Fi mesh topology.
Figure 7: Phase 2 WiMAX as an intra-mesh backhaul option.
For intra-mesh connectivity, Wi-Fi offers advantages today.
Industry-proven Wi-Fi chipsets and radios are readily available
and economical. They readily operate in unlicensed regions of
the spectrum. The result is an intra-mesh technology that
offers reliable high performance at the lowest cost. Essentially,
the inter-mesh backhaul connections can reduce costs
associated with wiring each node. When they become
available, the dual Wi-Fi and WiMAX APs will provide higher
performance and an even more robust solution. As shown in
Figure 8, APs with dual Wi-Fi and WiMAX radios can be easily
integrated into a mesh network. The solid blue links show
WiMAX backhaul and inter-mesh connectivity.
Figure 8: Phase 3 WiMAX as a client connection option.
With the emergence of WiMAX in the near future,
deployments that combine the two technologies can be
constructed to take advantage of the strengths of both Wi-Fi
and WiMAX.
Figure 6 shows the topology that could be used by a
municipality that wants to extend broadband connectivity to
two new rural community centers and a park. The
municipality wants to provide free Internet service to local
residences and staff to promote education, cultural arts and
local businesses. The deployment must be completed within
three months. A combination of WiMAX and Wi-Fi mesh-
network topology provides the best solution for this
13
14
situation. WiMAX can be used to aggregate the community
centers. WiMAX extends the reach of broadband, while the
proprietary Wi-Fi mesh network available today can provide
mobile client access throughout the community centers and
park. As dual-mode Wi-Fi and WiMAX cells are introduced
into high-capacity network centers in licensed or unlicensed
bands. The WiMAX cells will interoperate seamlessly with
existing Wi-Fi cells; always selecting the best path for
delivering maximum user throughput end-to-end.
Deployment Examples
Deployments of last-mile wireless solutions during 2004 and 2005 require analysis of immediate service needs and technology
investment goals. Here are some examples of how these needs can be balanced.
WiMAX Recommended
Last Mile
Challenge Solution
A WISP wants to expand its service coverage to underserved markets. WiMAX provides the best and the most cost-effective broadband
QoS is a signicant factor for this deployment because some of the solution to this challenge because the cost of deploying and providing
new customers are local government and small and medium traditional broadband services is prohibitively expensive. WiMAX is
businesses requiring a guaranteed level of service for certain designed from the ground up to provide a fast, cost-effective and
applications. Deployment cost and vendor interoperability is key easy-to-deploy solution with built-in QoS. WiMAX is based on
because many users within the target-market segment may end up IEEE standards and WiMAX-certied products are vendor interoperable.
owning their own WiMAX CPE.
Backhaul
Challenge Solution
A carrier is deploying two new cell towers and a Wi-Fi hotspot in a WiMAX provides the best solution for this challenge because it
rural community within the next two months. They want to be able to provides a cost-effective, rapidly deployable point-to-point
connect their cell towers to their core network and the hotspot to backhaul solution.
the Internet.
Wi-Fi Mesh Recommended
Vertical (Government and Education)
Challenge Solution
A school with existing broadband access (based on T1) wants to Wi-Fi mesh-network topology provides the best solution to this
expand connectivity to a new a classroom building and the schools challenge because its a quick solution to expand coverage to an
main courtyard. Students and faculty members are mobile and use existing wireless infrastructure and because 802.11-based products
their notebook computers and PDAs to access the Internet and the are available today.
schools network resources. The primary use of the network is to
download documents and presentations, access Web-based portals
(such as a blackboard), review class schedules and send instant
messages to students and instructors. The solution is needed now
and the school is willing to work with a single vendor.
Key Terminology
3GPP Third Generation Partnership Project
AP Access point. An AP operates within a specific
frequency spectrum and uses an 802.11 standard specified
modulation technique. It informs the wireless clients of its
availability, authenticates and associates wireless clients to
the wireless network and coordinates the wireless clients use
of wired resources.
BS Base station
CSMA/CA Carrier sense multiple access with
collision avoidance
CSMA/CD Carrier sense multiple access with collision
detection
DOCSIS Data over Cable Service Interface Specification
DSL Digital subscriber line
DSSS Direct sequence spread spectrum
ETSI European Telecommunications Standards Institute
FCC Federal Communications Commission
IEEE Institute of Electrical and Electronics Engineers
IP Internet Protocol
LAN Local area network
MAC address Media access control address. This address
is a computers unique hardware number.
MAN Metropolitan area network
OFDM Orthogonal frequency division multiplexing
OFDMA Orthogonal frequency division-multiple access
P2P Point-to-point
P2MP Point-to-multi-point
PAN Personal area network
PHY Physical layer
PoP Point of presence
QoS Quality of service
RF Radio frequency
SS Subscriber station
UWB Ultra-wide band
VoIP Voice over Internet Protocol
WAN Wide area network
Wi-Fi Wireless fidelity. Used generically when referring to
any type of 802.11 network, whether 802.11b, 802.11a,
dual-band, and so on.
WiMAX Worldwide Interoperability for Microwave Access
WISP Wireless Internet service provider
WLAN Wireless local area network
WMAN Wireless metropolitan area network
WWAN Wireless wide area networks
15
This document and related materials and information are provided as is with no warranties, express or implied, including but not limited to any implied
warranty of merchantability, tness for a particular purpose, non-infringement of intellectual property rights, or any warranty otherwise arising out of any
proposal, specication, or sample. Intel assumes no responsibility for any errors contained in this document and has no liabilities or obligations for any
damages arising from or in connection with the use of this document.
Intel, Intel Centrino and the Intel logo are trademarks or registered trademarks of the Intel Corporation or its subsidiaries in the United Sates and other countries.
*Other names and brands may be claimed as the property of others.
Copyright 2004 Intel Corporation. All rights reserved. 1004/JR/MESH Please Recycle 304471-001US
Additional Resources
White Papers
Intel Corp., IEEE 802.16* and WiMAX: Broadband Wireless
Access for Everyone, 2003,
www.intel.com/ebusiness/pdf/wireless/intel/80216_wimax.pdf
WiMAX Forum, WiMAXs Technical Advantage for Coverage
in LOS and NLOS Conditions, Aug. 2004,
www.wimaxforum.org/news/downloads/
WiMAXNLOSgeneral-versionaug04.pdf
General Resources
WiMAX Forum: www.wimaxforum.org
WiMAX technology overview:
www.intel.com/netcomms/technologies/wimax
WiMAX World Conference & Exposition:
www.wimaxworld.com
Tropos Networks: www.tropos.com